How Content Delivery Networks (CDNs) Work and Why They Make the Web Fast

The Problem CDNs Solve

When a user in Tokyo requests a webpage hosted on a server in Virginia, the data must travel roughly 11,000 kilometers each way. At the speed of light through fiber-optic cable, that trip takes about 70 milliseconds. Add protocol overhead, server processing time, and multiple round trips for a modern webpage, and load times can easily exceed a second — long enough to drive users away and hurt search engine rankings.

A Content Delivery Network (CDN) solves this by distributing copies of your content to servers located all over the world. When a Tokyo user requests your page, they get it from a nearby server in Singapore or Tokyo itself — reducing the round-trip distance from thousands of kilometers to perhaps a few hundred.

What a CDN Actually Is

A CDN is a geographically distributed network of proxy servers called Points of Presence (PoPs) or edge servers. Each PoP stores cached copies of content from an origin server — the original source of truth.

Major CDN providers include Cloudflare, Akamai, Amazon CloudFront, Fastly, and Google Cloud CDN. The largest CDNs have hundreds of PoPs in cities around the world. Cloudflare, for example, operates in over 310 cities across 120 countries.

How Request Routing Works

When a user types a URL or clicks a link, several steps determine where their request is served from:

1. DNS resolution — the CDN operator takes control of the domain's DNS records. Instead of resolving your domain to a single origin IP address, the CDN's DNS returns the IP of the nearest edge server.
2. Anycast routing — many CDNs use anycast, where the same IP address is announced from multiple locations. The internet's routing protocols automatically direct traffic to the closest (or least-congested) node that announces that address.
3. Cache check — the edge server checks whether it has a fresh, cached copy of the requested resource. A cache hit means it serves the content immediately; a cache miss means it fetches from the origin, caches the result, and then serves it.

Cache freshness is controlled by Cache-Control headers set by the origin server. Common directives include max-age (how many seconds until the cached copy expires) and s-maxage (a CDN-specific override).

What CDNs Cache

CDNs excel at caching static assets — content that is the same for every user:

• Images, videos, and audio files
• CSS stylesheets and JavaScript bundles
• HTML files for static websites
• Software downloads and firmware updates
• Streaming media segments (HLS/DASH chunks)

Dynamic content — personalized pages, shopping carts, API responses that vary by user — is harder to cache. Modern CDNs increasingly offer edge computing capabilities (Cloudflare Workers, Lambda@Edge) that let developers run code at the edge to dynamically assemble pages closer to the user while still avoiding round trips to the origin.

Performance Benefits Beyond Caching

CDNs improve performance through several mechanisms beyond simple geographic proximity:

• Persistent connections — edge servers maintain long-lived connections to the origin, reducing the overhead of establishing new connections for each request.
• Protocol optimization — many CDNs negotiate HTTP/2 or HTTP/3 between the user and edge, even if the origin only supports HTTP/1.1. These newer protocols reduce latency through multiplexing.
• TLS termination at the edge — the expensive TLS handshake happens at the nearby edge server, not the distant origin.
• Image optimization — some CDNs automatically compress and format images for the requesting device (WebP for browsers that support it, lower resolution for mobile).
• Minification — CDNs can strip whitespace and comments from CSS and JavaScript in real time.

Security Benefits of CDNs

CDNs have evolved from pure performance tools into security infrastructure:

• DDoS mitigation — a CDN's distributed capacity can absorb volumetric attacks that would overwhelm a single origin server. Cloudflare regularly mitigates attacks exceeding 1 Tbps.
• Web Application Firewall (WAF) — CDN-level WAFs inspect incoming requests for SQL injection, XSS, and other attack patterns before traffic reaches the origin.
• Bot management — CDNs can distinguish legitimate users from scrapers, credential-stuffing bots, and inventory hoarders.
• Origin protection — by hiding the origin server's real IP address behind CDN IPs, the origin is less exposed to direct attacks.

How CDN Pricing Works

Most CDNs charge based on a combination of bandwidth transferred (GB delivered to users) and requests made. Prices vary by region — bandwidth from North America and Europe tends to be cheaper than from Australia or South America. Many CDNs offer free tiers suitable for small sites: Cloudflare's free plan is generous enough for many small businesses.

For high-traffic media sites, CDN costs can run into thousands of dollars per month — but this is almost always offset by reduced origin server costs, improved conversion rates from faster load times, and avoided revenue loss from downtime.

CDNs and Modern Web Architecture

Modern frameworks like Next.js, Nuxt, and Astro are designed with CDN distribution in mind. The Jamstack architecture pre-builds entire websites into static files served entirely from CDN edge nodes, with no origin server involved at all for the majority of requests. Edge functions and edge databases (like Cloudflare D1 or Upstash) are extending this model to cover dynamic workloads as well.

Understanding how CDNs work is increasingly essential knowledge for web developers, because the choice of CDN — and how cache headers are configured — often has a larger impact on user experience than application-level optimizations.